Why “Open” is Better

Psychological models and tests influence decisions that can alter the course of people’s lives. They shape clinical assessments, workplace recruitment, educational support, research findings, insurance decisions and, increasingly, the software systems used to recommend, classify and predict human behaviour.

Because of that influence, the question of who owns and controls these tools is not merely a commercial matter. It is a question of scientific accountability, public trust and individual rights.

There is nothing inherently wrong with charging for professional services, publishing materials or protecting genuinely novel inventions. Developers need resources to conduct research, train practitioners, maintain assessment platforms and support responsible use. Some forms of intellectual-property protection can also provide temporary incentives for innovation.

However, psychological models and tests become problematic when commercial ownership gives a private organisation excessive control over the underlying concepts, scoring rules, evidence base, access conditions or public discussion. When a tool is used to make important judgments about people, secrecy and restricted access can conflict with the standards that psychology should uphold.

A more open approach—where models, measures, evidence and limitations are examined and improved by a broad range of qualified experts—offers important advantages. It does not mean that every test should be freely available without safeguards, or that expertise is unnecessary. It means that the knowledge underlying consequential decisions should be open to meaningful scrutiny and continued improvement.

Psychological tools are not ordinary consumer products

A personality questionnaire, cognitive assessment or clinical screening measure may look like a product that can simply be designed, packaged and sold. But such tools make claims about human beings. They may suggest that a person has a particular trait, condition, ability or risk profile.

That creates a higher standard of responsibility.

A typical consumer product can often be evaluated by whether it performs its advertised function. Psychological tests require more complicated questions:

• What exactly is being measured?
• Is the construct clearly defined?
• Does the test measure it consistently?
• Does it measure the same thing across cultures, languages and groups?
• Are the scores being interpreted appropriately?
• What are the consequences of false positives and false negatives?
• Does the test disadvantage particular populations?
• Is there evidence that the test predicts the outcome for which it is being used?
• Are users likely to mistake a probabilistic result for a fact about a person?

These are not questions that can be settled by branding, popularity or sales figures. They require continuing evaluation by researchers and practitioners who are able to inspect the assumptions, methods and evidence behind the tool.

Commercial control can make that process more difficult.

Ownership can restrict scientific scrutiny

Scientific knowledge develops through criticism, replication and comparison. A model becomes more credible when independent investigators can understand it, test it and challenge it. A test becomes more trustworthy when researchers can examine its items, scoring methods, validation studies and limitations.

Commercial ownership may restrict access to some or all of these materials. A publisher may limit the test questions to prevent copying. A company may keep scoring algorithms confidential. Researchers may be required to sign agreements that limit what they can disclose. Independent comparisons may be difficult if competing tests are not accessible on comparable terms.

Some restrictions have legitimate purposes. If test items are publicly available, people may rehearse them, compromising the usefulness of future assessments. Confidentiality can also protect sensitive content. Yet there is a difference between protecting a live test from contamination and preventing qualified researchers from evaluating its structure and evidence.

A responsible system could provide controlled access for independent review without making every item publicly searchable. For example, accredited researchers might inspect materials under conditions designed to prevent widespread disclosure. Statistical code, technical documentation and validation results could be published wherever doing so would not compromise test security. Retired items could be released for analysis after they are no longer in active use.

The principle should be proportionality: protect what genuinely needs protection, while making as much as possible available for scrutiny.

Proprietary control can turn uncertainty into authority

Psychological measurement is often presented with a level of certainty that the evidence does not justify. A neat score can appear objective even when the underlying construct is contested, the sample was narrow or the prediction is weak.

Commercial presentation can intensify this problem. A product must often be marketed clearly and confidently. Terms such as “accurate,” “scientifically validated” or “proven” may be used without sufficient explanation of what was actually validated and under which conditions.

Validation is not a single stamp of approval. It is a body of evidence gathered for particular interpretations and uses. A test may show reasonable reliability but poor predictive validity. It may work in one population but perform differently in another. It may support research classification but not individual diagnosis. It may be useful for generating a conversation while being unsuitable for high-stakes decisions.

Open development makes it harder for any one organisation to present a tool as more definitive than it is. When methods, findings and criticisms are visible, others can identify overstatements, test competing explanations and clarify where the tool should not be used.

This matters because labels influence how people are treated. A person classified as “high risk,” “low potential,” “resistant,” “unmotivated” or “unsuitable” may encounter barriers that persist long after the original assessment. The language surrounding a score can become more influential than the score’s actual statistical meaning.

Independent replication should not depend on commercial permission

Replication is central to reliable science. A finding that appears in one developer’s research programme may weaken when tested by independent teams, in different settings or with different samples.

When a psychological test is commercially controlled, replication may be expensive or procedurally difficult. Researchers may need to buy licences, obtain permission, complete training or accept restrictions on publication. These barriers can discourage independent work, especially among early-career researchers, small institutions and teams in lower-income countries.

The result may be a literature dominated by the test’s creators or commercial partners. That does not automatically make the evidence invalid, but it creates a risk of what might be called an evidence bottleneck: a small group controls both the tool and much of the research used to support it.

Open models reduce this bottleneck. Multiple teams can implement the model, identify weaknesses, propose alternatives and compare results. They can investigate whether findings hold across languages, cultures, occupations, age groups and clinical contexts.

This broader participation is especially valuable in psychology because human behaviour is shaped by social context. A model developed by a narrow group may reflect assumptions that are not obvious to its creators. Researchers with different experiences may notice missing variables, culturally specific interpretations or harmful consequences that others overlooked.

A model should be judged by evidence, not ownership

The commercial success of a model does not establish its scientific quality. A widely used assessment may be popular because it is easy to administer, familiar to organisations or supported by an effective sales network. Conversely, a well-supported open model may receive less attention because no company is promoting it.

When ownership and reputation become closely connected, users may confuse market presence with validity. Training providers may teach the tool because employers recognise its name. Consultants may prefer it because it offers a packaged framework. Organisations may adopt it because procurement systems favour established vendors.

An open ecosystem changes the basis of comparison. Tools can be assessed according to transparent criteria such as:

• quality and size of validation samples;
• reliability estimates and their limitations;
• evidence of predictive or construct validity;
• measurement invariance across relevant groups;
• rates of missing data and scoring errors;
• performance in realistic settings;
• replication by independent teams;
• accessibility and language availability;
• documented risks and appropriate-use boundaries;
• clarity of scoring and interpretation.

This does not eliminate disagreement. It makes disagreement more productive because participants can examine the same underlying evidence.

Open development can improve cultural and linguistic fairness

Many psychological tools are created in one language and cultural setting, then adapted for use elsewhere. Translation alone is not enough. Concepts can carry different meanings across cultures, and behaviours regarded as healthy, assertive, cooperative or concerning in one context may be interpreted differently in another.

Commercial tests may provide translated versions, but the process and evidence can be difficult for outsiders to inspect. There may be limited information about who conducted the adaptation, how items were selected, whether local experts were involved or whether scores have equivalent meaning across groups.

A more open process allows researchers from the relevant communities to participate directly. They can question assumptions, propose culturally appropriate items and examine whether the model captures local experiences rather than forcing them into categories developed elsewhere.

This is not an argument that every culture requires an entirely separate psychological science. It is an argument for wider participation in development and validation. Open collaboration can help distinguish features that generalise across settings from features that reflect a particular history, language or institutional environment.

Open models are easier to improve

No psychological model is final. Human behaviour is complex, theories change and new evidence can reveal weaknesses in established frameworks.

A closed commercial model may be updated slowly because changes create costs. Existing training materials, software, manuals and marketing claims may depend on the current version. The organisation may also have an incentive to preserve a recognisable brand, even when parts of the model need substantial revision.

Open models can develop through visible versioning. Researchers can document changes, explain why they were made and compare new results with earlier versions. Users can identify which version was used in a study or assessment. Problems can be logged rather than quietly handled through private support channels.

This kind of development resembles other areas of collaborative science and software. It supports an audit trail. People can see how the model changed, which evidence informed the change and whether the revision improved performance.

Open development does not guarantee improvement. Poor ideas can spread, contributors can disagree and governance can become difficult. But these are reasons to build strong review processes, not reasons to place all authority in one private organisation.

Public-interest infrastructure should remain available

Psychological models and tests are often used in areas with public consequences. Schools, hospitals, courts, employers and public agencies may rely on them. If essential assessment knowledge is controlled by a private vendor, institutions may become dependent on that vendor’s pricing, terms and technical decisions.

This can create several risks:

• licence fees may exclude smaller services;
• access may vary between wealthy and under-resourced regions;
• changes to the product may occur without broad consultation;
• organisations may struggle to migrate to alternatives;
• researchers may be unable to reproduce earlier results after a platform changes;
• people assessed by the system may have little opportunity to understand or challenge the result.

Public-interest tools should be treated as infrastructure rather than merely as branded products. This does not require every service to be free. It does suggest that core documentation, evidence and governance should remain accessible, and that multiple qualified providers should be able to implement the approach.

Public funding can support this work. Universities, professional bodies, charities and government agencies can cooperate to create openly documented measures. Commercial organisations can still provide hosting, training, implementation and support without owning the entire knowledge base.

Openness can support better informed consent

People who complete psychological tests are often told that their participation is confidential, but confidentiality is only one aspect of informed consent. Meaningful consent also requires a reasonable explanation of what is being measured, how results will be used, who will see them and what decisions may follow.

Opaque tools make these explanations harder. If the test’s scoring logic is secret, practitioners may be unable to answer basic questions. If the model is presented as proprietary intellectual property, the person assessed may have little practical way to challenge an interpretation.

Open documentation supports clearer communication. It allows practitioners to explain that a score is an estimate, not a complete description of a person; that measurement has uncertainty; and that results should be interpreted in context.

Openness also strengthens the right to contest. People should be able to ask for the purpose of an assessment, the evidence supporting its use and the process for correcting errors. In high-stakes settings, they should not be expected to accept an unexplained classification simply because a company owns the system that produced it.

Openness does not mean unrestricted public access

A common objection is that psychological tests must be protected. If the questions and answers are published, people may practise in advance, fake responses or manipulate results.

That concern is valid for some assessments, particularly when repeated administration and item security are important. But it does not justify total secrecy.

Several safeguards can coexist with openness:

1. Controlled researcher access: Independent reviewers can inspect materials under agreements that protect live items.
2. Release of retired items: Older items can be made available for research after they are no longer used operationally.
3. Open scoring code: Algorithms and statistical procedures can be published without exposing every current question.
4. Transparent technical manuals: Evidence, limitations, norms and appropriate uses can be documented in detail.
5. Independent governance: A committee representing researchers, practitioners and affected communities can oversee updates.
6. Version control: Changes can be recorded so that results from different versions remain interpretable.
7. Professional training: Access to administration and interpretation can require demonstrated competence.
8. Use restrictions: Open licensing can prohibit harmful or misleading uses while permitting research and legitimate practice.

The goal is not to remove all boundaries. It is to ensure that boundaries protect assessment quality and personal information rather than functioning primarily as barriers to criticism.

Open models require responsibility and governance

There is a risk of treating “open” as automatically synonymous with “good.” It is not. An openly published model can be poorly designed, biased, copied without understanding or used beyond its evidence base.

Open development therefore needs standards. Contributors should disclose conflicts of interest. Studies should report methods clearly. Data should be handled ethically. Changes should be reviewed. Users should receive guidance about uncertainty and limitations. The communities affected by a model should have a meaningful voice in its development.

Licensing also matters. An open model should state what others may do with it and what forms of attribution are required. It should discourage deceptive marketing and unsupported clinical claims. Where personal data are involved, openness must not mean exposing identifiable information.

The strongest approach combines openness with professional accountability. Transparency makes evaluation possible; governance helps turn evaluation into responsible practice.

Commercial organisations can still make valuable contributions

Criticising commercial control is not the same as criticising commercial participation. Companies can contribute substantially to psychological assessment. They may fund large studies, build reliable platforms, create accessible interfaces, provide practitioner training and support implementation in organisations that lack internal expertise.

The issue is whether commercial participation must include exclusive control over the model and its evidence.

A healthier arrangement could separate different layers:

• the underlying theoretical model may be openly documented;
• core measures may be available under a fair licence;
• implementation software may be commercially provided;
• specialised support and training may be paid services;
• sensitive live test materials may receive controlled protection;
• independent researchers may have access for validation;
• users may retain the freedom to compare alternatives.

This allows businesses to earn revenue from genuine value without making public knowledge dependent on a single owner.

The practical standard should be proportional to the consequences

Not every questionnaire requires the same level of openness. A private self-reflection exercise used for personal interest is not equivalent to a test used to diagnose a mental-health condition, determine employment suitability or influence legal decisions.

The more consequential the use, the stronger the case for:

• transparent evidence;
• independent validation;
• public documentation;
• clear error rates;
• examination of group differences;
• accessible challenge procedures;
• oversight by qualified professionals;
• limits on automated interpretation;
• regular review and withdrawal when evidence is inadequate.

A tool used for low-stakes reflection may reasonably have fewer formal requirements. A tool used to restrict opportunities or assign clinical labels should meet a much higher threshold.

This principle also helps avoid an unproductive argument between total secrecy and total openness. The real question is: what information must be available to whom, and under what safeguards, for this use to be scientifically and ethically defensible?

A more open future for psychological science

Psychology benefits when models are treated as provisional explanations rather than finished products. They should be tested against evidence, revised when necessary and compared with alternatives.

Commercial ownership can be compatible with that process, but exclusive control creates structural pressures toward opacity, brand protection and dependence. When private organisations control the language, measures and scoring systems used to describe people, their interests may become difficult to distinguish from scientific judgment.

More open models offer a better foundation. They invite criticism, support replication, involve a wider range of expertise and make cultural assumptions easier to identify. They can reduce costs, expand access and allow practitioners to select tools based on evidence rather than marketing. They can also give assessed individuals a stronger basis for understanding and questioning decisions that affect them.

The answer is not to abandon professional expertise, intellectual property or responsible test security. It is to place those elements within a broader public framework. Psychological knowledge should be developed with enough openness that qualified people can inspect it, affected communities can contribute to it and independent researchers can improve it.

When a psychological model influences opportunities, treatment or social status, no single company should be treated as the final authority over what the model means or whether it works. The stronger principle is shared development, transparent evidence and accountable use.

That approach is not perfect, but it is more consistent with the central values of science: scrutiny, correction, participation and respect for the people whose lives are being measured.

Sources

1. American Educational Research Association, American Psychological Association, & National Council on Measurement in Education. Standards for Educational and Psychological Testing. 2014.
2. American Psychological Association. Ethical Principles of Psychologists and Code of Conduct, including standards concerning assessment, informed consent, test construction and interpretation. Current version accessed for general ethical principles.
3. International Test Commission. The ITC Guidelines for Test Use. International Test Commission.
4. International Test Commission. The ITC Guidelines for Translating and Adapting Tests. International Test Commission.
5. World Health Organization. Ethics and Governance of Artificial Intelligence for Health. 2021.
6. National Academies of Sciences, Engineering, and Medicine. Reproducibility and Replicability in Science. National Academies Press, 2019.
7. Nosek, B. A., et al. “Promoting an Open Research Culture.” Science, vol. 348, no. 6242, 2015, pp. 1422–1425.
8. Open Science Collaboration. “Estimating the Reproducibility of Psychological Science.” Science, vol. 349, no. 6251, 2015.
9. Flake, J. K., and E. M. Fried. “Measurement Schmeasurement: Questionable Measurement Practices and How to Avoid Them.” Advances in Methods and Practices in Psychological Science, vol. 3, no. 4, 2020, pp. 456–465.
10. American Psychological Association. Multicultural Guidelines: An Ecological Approach to Context, Identity, and Intersectionality. 2017.

Rob Perin, CCHT

Hypnotechs Hypnotherapy & Hypnosis